Matlab实现图像数据集的特征提取与分类器设计_掌握几种常用的局部特征点图像特征提取算法,构建一个图像分类器,并在matlab环境下

包含内容：

1. 特征：LBP、HOG、Gabor、GLCM(灰度共生矩阵)、Hu(Hu矩)；
2. 分类器：KNN(K-近邻)、SVM(支持向量机)、NB(朴素贝叶斯)、Tree(决策树)。

数据集读入函数代码：

function [trainData,trainLabels,testData,testLabels] = ReadDataSet(fileRoot,imageFormat,trainNum,testNum)
%读取数据集
dataSetFiles = dir(strcat(fileRoot,'\s*'));
classNum = length(dataSetFiles);
trainLabels = [];
testLabels = [];
for i = 1 : classNum
    imageData = dir(strcat(fileRoot,'\',dataSetFiles(i).name,'\*.',imageFormat));
    for j = trainNum(1) : trainNum(2)
        image = im2double(imread(strcat(fileRoot,'\',dataSetFiles(i).name,'\',imageData(j).name)));
        trainData(:,:,(trainNum(2)-trainNum(1)+1)*(i-1)+j-trainNum(1)+1) = image;
        trainLabels = [trainLabels;i];
    end
    for j = testNum(1) : testNum(2)
        image = im2double(imread(strcat(fileRoot,'\',dataSetFiles(i).name,'\',imageData(j).name)));
        testData(:,:,(testNum(2)-testNum(1)+1)*(i-1)+j-testNum(1)+1) = image;
        testLabels = [testLabels;i];
    end
end
end
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

特征提取函数代码：

1. LBP

function total_LBP_Features = Extract_LBP_Features(data)
%提取LBP特征
[~,~,totalNum] = size(data);
total_LBP_Features = [];
for i = 1 : totalNum
    LBP_Features = extractLBPFeatures(data(:,:,i),'CellSize',[16,16],'Upright',true);
    numNeighbors = 8;
    numBins = numNeighbors*(numNeighbors-1)+3;
    LBP_CellHists = reshape(LBP_Features,numBins,[]);
    LBP_CellHists = bsxfun(@rdivide,LBP_CellHists,sum(LBP_CellHists));
    LBP_Features = reshape(LBP_CellHists,1,[]);
    total_LBP_Features = [total_LBP_Features;LBP_Features];
end
end
1
2
3
4
5
6
7
8
9
10
11
12
13
14

2. HOG

function total_HOG_Features = Extract_HOG_Features(data)
%提取HOG特征
[~,~,totalNum] = size(data);
total_HOG_Features = [];
for i = 1 : totalNum
    total_HOG_Features = [total_HOG_Features;extractHOGFeatures(data(:,:,i))];
end
end
1
2
3
4
5
6
7
8

3. Gabor

function total_Gabor_Features = Extract_Gabor_Features(data)
%提取Gabor特征
[~,~,totalNum] = size(data);
total_Gabor_Features = [];
GaborArray = gabor(10,[0,45,90,135]);
for i = 1 : totalNum
    total_Gabor_Features =<
1
2
3
4
5
6